Re: *IT WORKS* Re: CardBus USB 2.0 Controller (NEC uPD)

2003-06-14 Thread M. Warner Losh
In message: [EMAIL PROTECTED] Craig Boston [EMAIL PROTECTED] writes: : On Fri, 2003-06-13 at 14:17, Tony Naggs wrote: : Yes, I think you should also do this for Uhci. There are probably not : many straight Uhci USB 1.1 Cardbus cards, but it is likely some of the : USB 2.0 cards

Re: *IT WORKS* Re: CardBus USB 2.0 Controller (NEC uPD)

2003-06-13 Thread Tony Naggs
In message [EMAIL PROTECTED], Craig Boston [EMAIL PROTECTED] wrote On Thu, 2003-06-12 at 08:14, Anthony Naggs wrote: Sorry, I don't understand this comment. All OHCI, UHCI EHCI USB controllers need PCI bus mastering in order to read update their various lists of pending completed

Re: *IT WORKS* Re: CardBus USB 2.0 Controller (NEC uPD)

2003-06-13 Thread Craig Boston
On Fri, 2003-06-13 at 14:17, Tony Naggs wrote: Yes, I think you should also do this for Uhci. There are probably not many straight Uhci USB 1.1 Cardbus cards, but it is likely some of the USB 2.0 cards have an UHCI controller rather than OHCI for USB 1.x support. Never mind, it seems Warner

Re: *IT WORKS* Re: CardBus USB 2.0 Controller (NEC uPD)

2003-06-12 Thread John-Mark Gurney
Craig Boston wrote this message on Wed, Jun 11, 2003 at 22:44 -0500: Believe it or not, after futzing with the debugger for hours, reading the OHCI spec, and trying to figure out why writing to the control registers works exactly as it should but the card seems to ignore the ED list, I

Re: *IT WORKS* Re: CardBus USB 2.0 Controller (NEC uPD)

2003-06-12 Thread Bernd Walter
On Thu, Jun 12, 2003 at 01:10:24AM -0700, John-Mark Gurney wrote: Craig Boston wrote this message on Wed, Jun 11, 2003 at 22:44 -0500: Believe it or not, after futzing with the debugger for hours, reading the OHCI spec, and trying to figure out why writing to the control registers works

Re: *IT WORKS* Re: CardBus USB 2.0 Controller (NEC uPD)

2003-06-12 Thread Craig Boston
On Thu, 2003-06-12 at 03:10, John-Mark Gurney wrote: Craig Boston wrote this message on Wed, Jun 11, 2003 at 22:44 -0500: Hey, thanks for the great work. This got me past the same problem on the sparc box I have... Don't look at me -- thank Warner for all his hard work on Cardbus and the

Re: *IT WORKS* Re: CardBus USB 2.0 Controller (NEC uPD)

2003-06-12 Thread Anthony Naggs
In article [EMAIL PROTECTED], Craig Boston [EMAIL PROTECTED] writes Believe it or not, after futzing with the debugger for hours, reading the OHCI spec, and trying to figure out why writing to the control registers works exactly as it should but the card seems to ignore the ED list, I decided to

Re: *IT WORKS* Re: CardBus USB 2.0 Controller (NEC uPD)

2003-06-12 Thread Craig Boston
On Thu, 2003-06-12 at 08:14, Anthony Naggs wrote: Sorry, I don't understand this comment. All OHCI, UHCI EHCI USB controllers need PCI bus mastering in order to read update their various lists of pending completed transfers. That was speculation on my part as to why bus mastering was not

Re: *IT WORKS* Re: CardBus USB 2.0 Controller (NEC uPD)

2003-06-12 Thread John-Mark Gurney
Bernd Walter wrote this message on Thu, Jun 12, 2003 at 14:30 +0200: On Thu, Jun 12, 2003 at 01:10:24AM -0700, John-Mark Gurney wrote: Craig Boston wrote this message on Wed, Jun 11, 2003 at 22:44 -0500: pci_enable_busmaster(self); near the top of ohci_attach() in ohci_pci.c

*IT WORKS* Re: CardBus USB 2.0 Controller (NEC uPD)

2003-06-11 Thread Craig Boston
Believe it or not, after futzing with the debugger for hours, reading the OHCI spec, and trying to figure out why writing to the control registers works exactly as it should but the card seems to ignore the ED list, I decided to try something completely crazy and put the line

Re: *IT WORKS* Re: CardBus USB 2.0 Controller (NEC uPD)

2003-06-11 Thread Craig Boston
Cut-and-paste of the patch since the attachment disappeared... Probably won't apply cleanly because of tabs. --- ohci_pci.c.orig 2003-06-11 22:32:42.0 -0500 +++ ohci_pci.c 2003-06-11 22:01:43.0 -0500 @@ -173,6 +173,8 @@ /* XXX where does it say so in the spec? */